Clinical Pharmacology Review NDA 22-307, Prasugrel

3 DETAILED LABELING RECOMMENDATIONS

GENERAL

6120/2008

The Agency considered that the overall infonnation regarding Clinical Phannacology providedin the original NDA 22-307 was appropriate.

CLINICAL PHARMACOLOGYLABELING COMMENTS

Highlights Section:

r

Page 55 of262

b(4)

b(5)

f Page(s) Withheld

__ Trade Secret / Confidential (b4)

_~X_ Draft Labeling (b4)

X Draft Labeling (b5)

__ Deliberative Process (b5)

Cliniul Pharmacology Reviel\' NDA 22-307, Prasugrel

4 APPENDIX II:

4.1 Individual In Vitro Study Reviews

6/20/2008

4.1.1 In Vitro Protein Binding of Metabolites OF CS-747 (Report No. ATR-151-053)

Investigator: Atsushi Kurihara, Ph.D., Drug Metabolism and Pharmacokinetics ResearchLaboratories, Sankyo Co., Ltd., 2-58 Hiromachi l-chome, Shinagawa-ku, Tokyo 140-8710,JapanDate of completion: June 27 2005

Objectives

Methods

Assay

To determine the extent of the protein binding ratios of inactive metabolitesofCS-747, i.e., R-95913, R-I06583, R-I00932, R-119251 in plasma and R­138727 in 4% Human serum albumin (HAS)Unbound fractions of the main metabolites of CS-747 in plasma, R-95913,R-119251, R-I00932 and R-I06583, were measured by anultracentrigufation method.Rat, dog and human plasma at the concentrations of 50, 100 and 500 ng/mLor 100, 500 and 1,000 ng/mLEach mixture (fmal volume: 1 mL) was incubated at 37°C for 5 min. Then,centrifuged and ultracentrifuged at 436,000 g for 140 min at 15°CThe protein-containing fraction (lower layer), protein-free fraction (middlelayer) and lipoprotein-containing fraction (upper layer) separated. ,.-- t

------j .- (for R-95913, R-I00932, R--106583 and R-138727 assay) or - - (for R-II9251assay) as the internal standard. The mixtures were vortexed and centrifugatedat 14,000 rpm for 3 min at 4°C.The protein binding ratio in rat, dog and human plasma was calculated. HSAsolution at 4% was prepared with sodium phosphate buffer (pH 7.4). R­138727 was mixed with 4% HSA at concentrations of 50, 100 and 500ng/mL, and then each mixture (final volume: I mL) was incubated at 37°Cfor 5 min. After the incubation, the protein-free fraction and totalconcentration of R-138727 were measured as described above for the otherfour inactive metabolites.LC-MS for the total plasma concentrations of R-95913, R-I06583, R­100932, R-119251. Unbound fraction ofR-138727 (active metabolite ofCS­747) in 4% HSA was also measured by the ultracentrifugation method.

b(4)

ResultsThe protein binding ratios ofR-119251, R-I00932, R-I06583 and R-95913 are shown in Tables3, 4 and 5. The result of binding with HSA of R-138727 are also summarized in Table 6. Inplasma of each species, the protein binding ratios of R-95913, R-I06583 and R-I00932 weremore than 80% at concentrations of 50, 100 and 500 ng/mL. Binding ratio of R-l19251 was 71­77% in rat, 26-36% in dog and 76-77% in human plasma at 100, 500 and 1000 ng/mL. To

Page 60 of 262

Clinical Pharmacology Review NDA 22.307, Prasugrel 6/2012008

evaluate the plasma protein binding of the active metabolite, R-138727, commercially availablepurified HSA was used since this metabolite has been reported to be unstable in the plasma ofeach species.3) As a result, R-138727 was proved to be bound to 4% HSA with a binding ratio of98% at concentrations of 100 and 500 ng/mL.

Table 9. Protein binding ratios of R-95913, R-I00932, R-I06583 and R-119251 in rat (leftpanel) and dog (right panel) plasma

<R-95913>Prl)t~in binding ratio (%)50ng/mL lOOnglmL 500ng/mL

No. I 86.95 86.65 85.98No.2 89.20 87.52 84.84No.3 86.69 86.90 84.54Mean 87.61 87.02 85.12S.D. 1.38 0.45 0.76

<R-I00932>Protein binding ratio (%)50nglmL 100nll./mL 500ng/mL

No.1 93.57 92.14 92.13No.2 92.87 92.69 91.92No.3 92.15 92.61 92.\0Mean 92.86 92.48 92.05S.D. 0.71 0.30 0.11

<R·I06583>Prolc:in binding ratio (%)50nglmL 100 nglmL 500ng/mL

No.1 85.13 82.48 83.63No.2 82.84 82.81 82.30No.3 81.54 82.77 82.94Mean 83.17 82.69 82.96S.D. 1.82 0.18 0.67

<R-119251>Protein binding ratio (%)100nglmL 500ng/mL JOOOng/mL

No.1 68.78 76.82 76.39No.2 71.21 76.38 77.51No.3 73,06 75.80 77.60Mean 71.02 76.33 77.17S.D. 2.15 0.51 0.67

Page 61 of 262

<R-95913>Protein binding ratio (%)50ngimL lOOng/mL 500 nglmL

No.1 89.14 88.58 88.25No.2 93.07 93.27 92.53No.3 85.58 87.42 84.57Mean 89.26 89.76 88.45S.D. 3.75 3.10 3.98

<R-I00932>Proteinbinding roho (%)50rii/mL 100 ll!i7inL 500DWinL

No.1 84.94 85.17 79.35No.2 88.56 87.13 83.61No.3 81.89 77.86 76.69Mean 85.13 83.39 79.88SD. 3.34 4.89 3.49

<R-I06583>Protein binding ratio (%)50nglmL 100 ngllnL 500ngllnL

No.1 83.93 84.87 80.07No.2 86.51 84.74 82.24No.3 83.61 80.66 79.85Mean 84.68 83.42 80.72SD. 1.59 2:39 1.32

<R-1I9251>Protein binding ratio (%)l00ng/ml 500ng/01L JOOOuglmL

NO.1 32.46 28.12 26.53No.2 33.21 21.50 21.86No.3 40.95 31.37 30.70Mean 35.54 27.00 26.36S.D. 4.70 5.03 4.42

Clinical Pharmacology Review NDA 22-307, Prasugrel 6/20/2008

Table 10 Protein binding ratios of R-95913, R-100932, R-106583 and R-119251 in humanplasma

<R-95913>Protein binding ratio (%)50nglmL 100ng/mL 500ng/mL

No.1 93.60 93.72 94.19No.2 94.12 94.15 96.26No.3 95.42 93.70 96.31Mean 94.38 93.86 95.59S.D. 0.94 0.25 1.21

<R·I00932>Protein binding ralio (%)50nglmL 100ng/mL 500 nglmL

No.1 94.98 91.21 91.40No.2 93.45 89.51 89.70No.3 93.52 92.32 90.89Mean 93.98 91..01 90.66S.D. 0.86 1.42 0.87

<R-I06583>Protein binding ratio (%)50nglmL 100ng/mL 500nglmL

No.1 N.A. 96.79 95.28No.2 N.A. 94.15 94.26No.3 N.A. 95.88 94.27Mean N.A. 95.61 94.60S.D. N.A. 1.34 0.59

<R·1192~1>

Protein binding ratio (%)100 nglmL 500nglmL 1000 ngfmL

No.1 74.67 75.29 78.23No.2 76.54 75.74 74.89No.3 75.32 76.41 76.03No.4 78.50 77.07 78.29Mc:an 76.26 76.13 76.86S.D. 1.68 0.78 1.68

Table 11 Protein binding ratios ofR-138727 in 4% HSA

Protein binding ratio (%)50ng/mL 100nglml 500nglmL

No.1 N.A. 97.96 98.06No.2 N.A. 97.96 97.91No.3 N.A. 97.95 97.99Mean N.A. 97.96 97.99S.D. N.A. 0.01 0.08

Page 62 of262

Clinical Pharmacology Review NDA 22·307, PrasugreJ 6/2012008

4.1.2 Exploratory Studies on the in Vitro Biotransformation of Prasugrel (LY640315) byHuman Carboxylesterases 1 and 2 (Report # 2007IV-EIO02)'Reported by Eric T. WilliamsConducted by Nagy A. Farid, Everett J. Perkins, G. Douglas Ponsler, and Eric T. WilliamsEli Lilly and Company Lilly Corporate Center Indianapolis, IN 46285Date: October 2007

Objectives To examine in vitro the potential involvement of human carboxylesterases,hCEl and hCE2, the dominant forms in the liver and intestinal tract,respectively, in the bioactivation ofprasugrel.

Investigated Prasugrel, R-95913, and R-104731 (Sankyo Co., Ltd) The enzymes hCElCompound and hCE2, were expressed and purified by the Integrative Biology Protein

Expression group at Eli Lilly.Incubation at Prasugrel (in 2% acetonitrile) solutions at concentrations from 0.855 to 10937°C JlM pre-incubated (I min). The enzyme, hCEI (I Jlg/mL) or hCE2 (0.25

Jlg/mL) pre-incubated with buffer (5 min) and added to the prasugrel. AfterI, '2, 3, and 6 min reaction terminated by addition to 100 jlL of acetonitrilecontaining 2 Jlg/mL of R-104731, a d4-labeled R-95913, as the internalstandard.

Assay LC-MSIMSData Analyses Hydrolysis reaction rates (nmol of product/min/jlg of protein) were

calculated from the linear portion of the concentration versus time curve.Fittings: WinNonlin (pharsight Corp.; Mountain View, CAl. Michaelis-MeIiten, substrate inhibition, and Hill kinetic models with differingweighting were used.

Prasugrel is a prodrug that requires a two-stage biotransformation for the formation of the activemetabolite. The first step is the hydrolysis of the ester bond to yield R-95913 (Farid et al. 2007a)which may be the result of the carboxylesterases (CEs). This is followed by oxidation of R­95913 which has been shown to be mediated primarily by CYP3A and CYP2B6 (Rehmel et al.2006).The carboxylesterases are a multigene family that hydrolyze compounds containing an ester,amide, or thioester linkage. Two primary forms exist in humans, hCEl and hCE2.Carboxylesterases are broadly expressed throughout the body. Both forms have greatest mRNAexpression in the liver, while extrahepatic expression differs (Satoh et at 2002). For hCEI, theliver-dominant form, extrahepatic mRNA expression observed in decreasing order are thestomach, testis, kidney, spleen, and colon. The intestinal-dominant form, hCE2, has extrahepaticmRNA expression in decreasing order in the colon, small intestine, and heart.Although hCEI and hCE2 have overlapping subs.trate recognition, selectivity for substrates hasbeen observed (Satoh et al. 2002). Two products result from ester hydrolysis, an alcohol and anacyl moiety. In general, hCEl prefers substrates with a large acyl moiety, while hCE2 preferssubstrates with a large alcohol substituent. Based upon the known substrate recognition of thesetwo carboxylesterases, prasugrel would be predicted to be a preferred substrate for hCE2.This study aims to show the potential involvement of hCE1 and hCE2, the dominant forms in theliver and intestinal tract, respectively, in the bioactivation ofprasugrel.

Page 63 of262

Clinical Pharmacology Review NDA 22-307, Prasugrel 6120/2008

ResultsThe human carboxylesterases hCEI and hCE2 hydrolyzedprasugrel to R-95913. Figure 24(A)depicts the fit of standard Michaelis-Menten kinetics to the data for hCEI, with an apparent Km·of 9.25 ± 0.78 J.lM and an apparent Vmax of 0.725 ± 0.035 ornol ofR-95913/min/J.lg of protein.The results obtained using hCE2 were not fit to any standard kinetic model, as seen in (B). Thesponsor described the data using 2 models. The first data set (prasugrel concentrations of 0.855and 40.5 J.lM) was described with the Hill equation, K.s of 11.1 ± 2.8 J.lM, and Vmax of 19.0 ±2.8 ornol of R-95913/min/J.lg of protein, and N= 1.42 ± 0.12. The second portion of the curveresembles an inhibition plot and as such the data between 27.4 and 109 J.lM was modeled forinhibition to yield an apparent IC50 of76.5 ± 2.7 J.lM.

A

I··Km .;..25 ~ 0.70 ......Ym;IlC ·0.72&.t. D.GaS nmaI ofR-9S01~"cI prolofn

B

Figure 24. Formation of R-95913 by hCEI (A) and hCE2 (B). The points with the errorbars represent the average and standard error, while the lines represent the best model fitof their respective kinetic models. Inset shows the sigmoidicity of the Hill kinetics.

The hydrolysis of prasugrel with both hCE1 and hCE2 have similar Km and K.s values. Thehydrolysis activity (Vmax) of hCE2 was about 26-times higher than that of hCEI. This isconsistent with the published data regarding carboxylesterase substrate preferences (Satoh et al.2002).

Comment:

1. The raw data of this study were not available for the review.2. This study determined the ability of expressed and purified human carboxylesterases

I (hCE1) and 2 (hCE2), the dominant forms in the liver and intestine, respectively, tohydrolyze prasugrel to R-95913 in vitro. The study showed that these humancarboxylesterases efficiently catalyze the conversion ofprasugrel to R-95913, with 26times the rate of turnover by hCE2 compared to hCE1.

Page 64 of 262

Clinical Pharmacology Review NDA 22-307, Prasugrel 6/20/2008

4.1.3 Examination of Effects of LY640315 on CYPIA2 and CYP3A in Primary Culturesof Human Hepatocytes

Investigator: J.L. Fayer, Eli Lilly, Indianapolis, IN 46285, USAD t A t2005a e: ugusObjectives To examine the ability of R-95913 to induce the catalytic activities

associated with CYPIA2 and CYP3A in primary cultures of humanhepatocvtes from four separate donors.

Method of Primary cultures of human hepatocytes were treated for 72 hours with R-incubation 95913 at concentrations ranging from 0.1 to 100 JlM, and the effects of

treatment on catalytic activities associated with CYPIA2 and CYP3A werecompared to those activities in vehicle control cultures (0.1 % DMSO).Omeprazole and rifampicin were used to demonstrate induction responses forCYPIA2 and CYP3A activity. The cells were incubated with 100 mcMtestosterone or 2mcM 7-ethoxy-resorufin (with 3mM salicylamide) in HMMfor 30 minutes. Cells were harvested by scraping in 1.0 mL 100 mMpotassium phosphate buffer (pH 7.4), and were stored at approximately -80°C prior to determination ofprotein content.

Assay HPLC with UV detection: the Lilly Global Chromatography Data System(GCDS), version 2, Eli Lilly and Company,

50

200

150

~8'0100

j

100100.1

i

°-;-i--~---,.---__-~0,01

ResultsAt 0.1, 1.0, 10, and 100 JlM concentrations of R-95913, preparations in 3 human hepatocytes(HH) showed some changes in CYPIA2- mediated EROD activity, examples in 2 HH are shownbelow:

200 1150 ~

e "~- IE100 11l . Ia; I

Q. J50,

i

IR·95913] (11M) 0,01 0.1 10 100

The effect or R4a59,3 on etttoxyresorufin O-dHthytaseactivity fn humanhepa~preparation HH'994.

[R·95913j (liM)

Activit5es life expre$Sed " percem of control 1:: standarderror of the mean. Mun control activity (n. 3) was0.21.2 pmoVmWmg, and omeprROlltotrellftd cella hadm.toJd gJe-atar ltGIJvity Chan controls.

·"·value~ then 0.0$.

Flgure2: The effect of R-96913 on ethoxyresorufin o-deethylaseactivity In hum~n tlepatocyte preparation HH9'7.

Activities are expressed as percent of control :t: stan~derror of the mean. Mean control activity In : 3) was0.781 pmoUminlmg, and omperazofe·treated cells had 4·foldgreater actiVity than controls..

In the last HH (not shown), the differences were not statistically different from control activity atany of the concentrations examined.

Page 65 of 262

Clinical Pharmacology Review NDA 22-307, Prasugrel 6120/2008

All three preparations showed statistically significant increases in CYP3A-mediated activity.Donor preparation HH993 showed approximately 8-fold induction at 10 ~M R-95913 relative tocontrols, but activities returned toward controls at I00 ~M. Concentration-dependent inductionof up to about 6-fold was observed at all concentrations of R-95913 in preparation HH994.Concentration-related induction of up to about 3-fold was evident in preparation HH998 (notshown) at concentrations of 10 and 100 ~M.

1000)

1000.1 1 10

(R·9S913j <11M)

0.01

1000,

I) 750j ~r" j . " ----------:

250 1I ..--

o1 ,- --"-.....,..,"""""----..~ ......._-----, -y•••--.,,, ~""..

10010

{R·95913j (jIM)

0.10.01

750 ~II

I

j:1 .-Io-r-.-------r-----~'--_r__--.---.

The'~ct of R·95'" on tutosl«One GJi-hydroxybMacttvhy II) human hepatocyte pr~1Ion HHlt3.

FigUR 6: The effect of R-tS9t3 on testosterone ~-hydroxyfa"

acUvity In human hep.aloeyte preparation HH9t4.

Activitkos ... exp,...M<I .., percent of c.onttol t sbndatd.trw of U. ,"un. Mean control actMty en • 3) was18.3 pmoUmJn1mg. Dnd rifampJdn-ftaled ~Us had 25·Joktg.ellter activity than ttl. or cord.ot.

Activltles are expre5sed e perunt of control :t standarderrorofthc me-8n. Mean ~ontroillCttvtty 'n:;; 3) was209 pmoUmtnlmg,. and ri1ampicin.tfeated cells had ' ...foJdgreater activity (hln tbIIt of control.

Comment:This was a pilot study of the effects of catalytical activities for CYPlA2 and CYP3A in culturesofhuman hepatocytes. The results give some indication that the parent drug, prasugrel, may havea potential to induce hepatic CYP3A.

Page 66 of 262

Clinical Pharmacology Review NDA 22-307, Prasugrel 6/20/2008

4.1.4 In Vitro Interaction of LY640315 (CS-747) Metabolites R-95913, R-138727, and R­106583 with Human Cytochrome P450 CYP3A

Study 2002IV-D1002Conducted by: James A. Eckstein, Jessica L. Fayer, John B. Heim, Steve C. Kasper, Barbara J.Ring, and Steven A. WrightonEli Lilly and Company Lilly Corporate Center Indianapolis, IN 46285December 2002

Objective To evaluate the abilities of the prasugrel active metabolite R-138727 and twoinactive metabolites, R-95913 and R-106583, to inhibit metabolism mediatedbyCYP3A

Incubation Time: I-minuteConditions Assay of I'-hydroxy midazolam by LC/MS/MS.Midazolam: Human liver microsomes (0.1 mg protein) in 100 mM sodium phosphate

buffer (pH 7.4), 1 mM NADPH, and midazolam (2.5 llM) in the absence orpresence ofR-138727 or R-I06583 (0.5,2.5, 10,50, or 200 llM) or R-95913(10, 20, 40, or 60 J.lM)midazolam: 2.5, 5.0, 10,25, or 50 llM. Formation of I'-hydroxy midazolamunder these conditions was linear with respect to time.

Incubation Time: 10 minConditions Assay of 6B-hydroxy testosterone by HPLC. R-138727 was not examinedTestosterone: (co-elution of substances related to R-138727 with the 6B-hydroxy

testosterone).Human liver microsomes (0.075 mg protein) in 100 mM sodium phosphatebuffer (pH 7.4), 1 roM NADPH, and testosterone (5.0 J!M) in the absence orpresence of R-106583 (0.5, 2.5, 10, 50, or 200 J!M) or R-95913 (25, 100,150, or 200 J.lM), and testosterone (5.0, 25, 50, 150, or 250 J!M). Formationof 6B-hydroxy testosterone under these conditions was linear with respect totime.

Data analysis WinNonlin Professional, version 3.1 was used to estimate parameters: Km,Vmax, and Ki and the standard errors

Results

The formation of l'-hydroxy midazolam by R-95913 was described with a competitive inhibitionmodel withKm 4.5 ± 0.7 llMVmax 921 ± 48 pmol/min/mg protein andKi 12.6 ± 2.2 llM.There was::::; 11% inhibition ofCYP3A-mediated midazolam 1'-hydroxylation by R-138727 (seebelow)

Page 67 of262

Clinical Pharmacology Review NDA 22.307, Prasugrel 6/20/2008

Table 12. Effect of R-138727 In Vitro on the CVP3AMediated Metabolism of Midazolamto l'-Hydroxy Midazolam

Concentration ofR-B8727(11M)

o0.52.51050200

l' -Hydroxy Midazolam FormationPerceut of Control(pmol/millimg)

348 100343 99338 97368 106334 96311 89

and < 18% inhibition of CYP3A-mediated mjdazolam 1'-hydroxylation by R-I06583 (seebelow).

Table 13. Effect of R-106583 In Vitro on the CYP3A Mediated Metabolism of Midazolamto 1'-Hydroxy Midazolam

Concentration ofR-106583(!-1M)

o0.52.51050200

1'-Hydroxy Midazolam Formation(prool/minimg)

348355328346286312

Percent of Control

100102941008290

The fonnation of 6B-hydroxy testosterone was described with a simple Michaelis-Mentenkinetics by the human microsomal mixture in the inhibition study with R-95913, yielding anapparent KIn valu.e of 42.5 ± 5.0 IlM and Vmax of2259 ±88 pmollmin/mg protein.The inhibition of testosterone metabolism by R-95913 was found to model best tononcompetitive inhibition yielding an apparent Ki value of 55.9 ± 4.2 IlM. There was:=:; 24%inhibition ofCYP3A-mediated testosterone 6J3-hydroxylase activity by R-I06583 (see below).

Table 14. Effect ofR-106583 In Vitro on the CVP3A Mediated Metabolism of Testosteroneto 68-Hydroxy Testosterone

Concentration ofR-106583(pM)

o0.52.51050200

Comments:

613-Hydroll.)' Testosterone Formation(pmol/minimg)

88.370.780.887.884.667.1

Percent ofControl

10080921009676

1. The predicted in vivo inhibition by R-95913 of the CYP3A was calculated based onthe highest reported mean Cmax value for R-95913 as 616 nglmL (1.9 1lM) followinga 60 mg loading dose ofprasugrel. An apparent Ki value for R-95913 was estimated